INDEX

Explanations

bullet points and lists

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 délic

1.54

 tùy

1.48

lust

1.45

 padrões

1.44

 నేపథ

1.43

étera

1.42

 antigos

1.42

 nifty

1.38

ارے

1.37

 frivolous

1.34

POSITIVE LOGITS

 refusal

1.30

째

1.28

찬

1.24

фі

1.22

入れ

1.20

 denials

1.20

 involvement

1.18

handedly

1.16

 excluded

1.16

exclude

1.15

Activations Density 0.682%